1 14 GEOLOGY OF THE HIGH PLATEAUS. 



a, general theory of volcanism, for neither the facts nor the antecedent gen- 

 eralizations are ready for it. Such a theory must be the -work of several 

 generations to come, and must gradually grow into form and coherence as 

 all great theories have done heretofore. Yet there are a few conceptions 

 of a high degree of generality which, perhaps, contain the germs of a theory, 

 though in their present condition they are vague and formless. They may 

 be said to resemble stones in the quarry, rough and unhewn, but which may 

 some time become corner-stones, columns, and entablatures in the future edi- 

 fice. I shall propose some of these considerations, not in the form of a con- 

 nected theory of volcanism, but as partial constituents of a theory in a 

 highly generalized form, taking care to proceed no further than existing 

 knowledge may afford at least some justification in proceeding. 



I. The first consideration has reference to the probable subterranean 

 locus of volcanic activity. In the present stage of our knowledge it 

 seems little credible that the sources of eruptive materials can be located 

 at very great depths. It is almost impossible that they could have 

 emanated from a general liquid interior. Taking the common notion that 

 the earth has formed, by cooling, an external rocky shell, enveloping a 

 nucleus which was once an intensely heated liquid, and which may still be 

 so, either partially or wholly, the ordinary principles of hydrostatics lead us 

 to conclude that all the primordial volcanic energy ought to have been 

 exhausted even before a stable crust could have been first formed. We 

 are in the habit of regarding the earth as hot within, but gradually dissi})at- 

 ing its heat by conduction through the crust and by radiation into space, 

 and if this conception have any truth, or even verisimilitude, then the erup- 

 tion of portions of its primordial liquid masses ought to become more and 

 more difiicult with the process of ages — nay, ought to have ceased at a 

 period long anterior to the most ancient of any of which systematic geology 

 can take direct cognizance ; for secular cooling can only strengthen the 

 rigid envelope and continually abstract from the heated magmas below the 

 heat which renders them liquid and eruptib^ We cannot in this connec- 

 tion ignore the plainest consequences of hyvlrostatic laws. A solid crust 

 covering a fluid nucleus, or a portion of that crust covering a large liquid 

 vesicle, could not remain stable for an hour unless the liquid were denser 



